Many buildings use metal detectors to screen incoming visitors for weapons such as firearms or knives. These detectors are most commonly seen at airports, but the use of metal detectors in schools and courthouses is increasing due to rising concern about the possibility of weapons being brought into these institutions.
A common problem with conventional metal detectors is that they cannot distinguish between objects containing iron ("ferrous") and those that do not contain iron ("non-ferrous") The vast majority of weapons are manufactured from ferrous metals, but since many metal detectors will detect non-ferrous metals as well as ferrous ones, the detectors will sound an alarm in the presence of many common metallic objects, such as keys, belt buckles and coins, in addition to weapons. Since most people entering a building are likely to be carrying keys or coins, conventional metal detectors tend to sound an alarm for nearly every person walking through it, requiring security officers to either ask the person to remove the keys or coins from his or her pockets or use a separate metal detecting wand to pinpoint the location of the offending metal objects. Both alternatives tend to be time-consuming and somewhat of an inconvenience, particularly if there are many people waiting in line to pass through the metal detector.
Most metal detectors use a "beat frequency oscillator" ("BFO") detector. This type of detector has a search loop oscillator and reference oscillator. These two oscillators oscillate at the same frequency if there is no metal near the detector. When a metallic object passes through the BFO detector, the search loop oscillator frequency changes in relation to the frequency of the reference oscillator. This causes the beat frequency between the search loop oscillator and the reference oscillator to increase, setting off an alarm. The main problem with the BFO detector is that it will detect metallic materials indiscriminately; it does not distinguish between ferrous metals (which are most commonly used in weapons) and other metallic substances, such as those used in coins and keys.
One possible method of distinguishing between ferrous and non-ferrous metals is by using a magnetometer, such as the one disclosed in U.S. Pat. No. 5,432,445 issued to Dinsmore et al. Since only ferrous metals are magnetic, magnetometers can distinguish between ferrous and non-ferrous metals. The sensor shown in Dinsmore et al. requires an outside coil wrapped around a pair of closely matched coils for sensing the magnetic field. Each of the matched coils has a magnetically permeable strip in its center. The matched coils must be identical and the magnetic characteristics of the center strips must also be identical, or a "mirror image" of each other.
Manufacturing of the center strips is particularly cumbersome and expensive; since each of the strips is punched from a single sheet of metal, the mechanical stresses from the punching process changes the magnetic characteristics of every strip. The strips must then be heat-treated through a hydrogen annealing method in an attempt to bring the strips'magnetic properties back to their original state. However, it is impossible to bring the magnetic properties of each individual strip to exactly the same state as before the punching process or to match the magnetic properties between strips. Since the matched coils and center strips must be "mirror images" of each other, any mismatch in the characteristics between the coils and between the strips, no matter how slight, will render the sensor, and thus magnetometer, useless. In initial batches of sensors tested by the inventor, only about half of the finished sensors were usable.